Typically, an electrical component, such as an electrical connector, for example, may include a plurality of electrically-conductive contacts, the terminal portions of which may be arranged in a matrix of rows and columns, for example. The contacts L may be signal conductors or ground conductors, and may be arranged along columns in a signal-signal-ground arrangement. Adjacent signal contacts may form differential signal pairs, though the signal contacts may be single-ended signal conductors. Such a component may include any combination of differential signal pairs and single-ended signal conductors.
Terminal portions of the contacts may be received by a substrate, such as a backplane or printed circuit board. At the board termination level, the contact leads may be terminated in plated through-holes (PTHs). PTH technology is a method of PCB manufacture whereby traces on one layer may be electrically connected through previously-drilled through-holes to traces on another layer using a plating method. When PTH technology is employed, the footprint, or arrangement of through-holes, must align with the arrangement of the terminal portions of the leads so that the terminal portions of the leads may be received by corresponding through-holes in the substrate.
Plated through-holes, however, constrain routing density (i.e., the number of traces that may be disposed onto the surface of a board layer). Plated through-holes, therefore, tend to increase the number of board layers necessary to provide required routing. Increasing the number of layers, however, increases board thickness and manufacturing cost. It also increases capacitance due to the increased number of ground planes. Increasing capacitance decreases impedance. Consequently, the board impedance may be driven lower than the component impedance, which creates a undesirable discontinuity between the board impedance and the component impedance. It is desirable that the component impedance and the board impedance be matched as nearly as possible to avoid signal reflections that occur because of the impedance discontinuity. Such reflections created unwanted noise that degrades signal integrity.
Alternatively, an electrical component may be mounted to a circuit board using surface mount technology (SMT). SMT involves electrically connecting terminal ends of the contacts to the surface of the substrate by electrically connecting each terminal end to a respective SMT pad located on the surface of the substrate. The terminal ends of the contacts, which may include electrically-conductive solder balls, for example, are typically soldered to the pads. On multi-layer boards, the SMT pads are typically electrically connected to vias that extend between the layers of the board and electrically connect SMT pads or traces on one layer to traces on another layer.
FIG. 1 depicts a typical SMT connector footprint comprising a plurality of SMT pads P arranged in a pad arrangement and a plurality of vias V arranged in a via arrangement. Each of the vias V is electrically connected to a respective one of the SMT pads P. The SMT pads P and vias V may be arranged in a so-called “dog-bone” pattern, as shown. As shown in the inset, a “dog bone” may include an SMT pad P, a via V, a via pad VP, and an electrically conductive via trace VT that electrically connects the via pad and the SMT pad. It should be understood, however, that vias and SMT pads need not be arranged in such a dog-bone pattern. Alternatively, an SMT pad may overlap partially or completely with a corresponding via pad such that there is a direct connection between the SMT pad and the via pad. Such a configuration is typically referred to as “via-in-pad.”
The SMT pads and vias may be arranged into rows and columns. As shown in FIG. 1, columns extend along the horizontal direction, perpendicular to the board edge E. Rows extend along the vertical direction, parallel to the board edge E. The spacing between the centerlines of adjacent rows may be referred to as the row pitch PR. The spacing between the centerlines of adjacent columns may be referred to as the column pitch PC.
The SMT pads P and vias V may be ground conductors or signal conductors. Signal conductors may be used in either single-ended or differential signal transmission. High-speed (i.e., greater than 1 GHz) connectors typically use differential signal pairs for signal transmission. In differential signal transmission, each signal conductor may be paired with an adjacent signal conductor. A respective ground conductor may be disposed between adjacent pairs of signal conductors. In some connector systems, ground conductors may be included to decrease cross-talk among the signal conductors, and to promote impedance-matching.
The pad arrangement depicted in FIG. 1 may be the same as the lead arrangement in the component to be surface-mounted onto the board. For example, the SMT pads may be arranged into rows and columns just as the terminal portions of the leads are arranged into rows and columns. Further, the row pitch PR and column pitch PC of the pad arrangement may be the same as the row pitch and column pitch of the lead arrangement.
Similarly, the via arrangement may be the same as the pad arrangement. That is, the vias V may be arranged into rows and columns, for example, just as the SMT pads P are arranged into rows and columns. Further, the row pitch PVR and column pitch PVC of the via arrangement may be the same as the row pitch PR and column pitch PC of the pad arrangement.
As described above, it may be desirable to decrease the number of board layers in an effort to reduce electrical discontinuity between the connector and the board and between different layers in the board. One way to accomplish this would be to increase the routing density on each layer of the board. Tying the via arrangement to the pad arrangement, however, tends to constrain the board designer's ability to improve routing density.
For example, in a typical connector having a 2 mm column pitch, a horizontal routing channel (i.e., the board space between adjacent columns) may be wide enough to include only one pair of traces. A typical connector having a 3 mm column pitch may be wide enough to include two pairs of traces. Similarly, in a typical connector having a row pitch of 1.4 mm, a vertical routing channel (i.e., the board space between adjacent rows) may be wide enough to include only a single trace. Consequently, in order to have a signal pair connected “vertically” out of the connector, a first trace must be disposed between a first pair of rows and a second trace must be disposed between a different pair of rows. It is usually desirable, however, to have the traces associated with a signal pair disposed as closely to each other as possible.